A power trading system, method and application thereof

By establishing a power supply and demand balance system for big data and artificial intelligence in the microgrid, predicting user power demand, allocating power and performing error compensation, the problems of supply and demand imbalance and prediction error in power transactions are solved, and transaction efficiency and economics are improved .

CN110717779BInactive Publication Date: 2024-11-12EVO ARTIFICIAL INTELLIGENCE TECH (JIANGSU) CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN201910824310.8
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2019-09-02
Publication Date
2024-11-12
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

In the existing power trading system, the supply and demand of electricity in the microgrid is unbalanced, the cost of energy storage devices is high, the trading mechanism is complex, and the lack of fair pricing strategies is lacking, and the power prediction errors are not effectively dealt with.

Method used

By establishing a power supply and demand balance system based on big data and artificial intelligence in the microgrid, users' power generation and load demand are predicted, and they are divided into power sellers and buyers' collections, establishing power purchase cost and seller benefits models, and determining power collection and distribution , realize the balance of power supply and demand, and adjust through prediction error compensation and time window.

Benefits of technology

It has achieved a stable balance between power supply and demand, increased the seller's power sales profit, reduced the buyer's power purchase cost and interference to the main power grid, and solved the economic losses caused by prediction errors.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN110717779B_ABST
    Figure CN110717779B_ABST
Patent Text Reader

Abstract

The present invention provides an electricity trading system, method and application thereof, wherein the electricity trading method comprises: predicting the power generation and load demand of each user in the microgrid in the next trading time window according to the preset time window of the microgrid power trading system, and dividing all users in the microgrid into a set of power sellers and a set of power buyers based on the power generation and load demand of each user obtained; establishing a power purchase cost model and a seller benefit model, collecting power from each power seller under the condition of minimizing the power purchase cost of the system dispatcher and maximizing the seller's benefit; establishing a buyer benefit model, allocating power to the power buyer under the condition of maximizing the sum of the benefits of all buyers. The present invention solves the problem of imbalanced power supply and demand for a single user by establishing a reliable and stable power trading system in a community microgrid, which not only improves the user's income but also reduces the interference to the main power grid.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of power trading, and in particular to a power trading system, method and application thereof. Background Art

[0002] Big data technology refers to the ability to quickly obtain valuable information from various types of data. Electricity is the practice of big data concepts, technologies and methods in the power industry, involving power generation, transmission, transformation, distribution, power consumption and dispatching. With the construction of smart grids and the application of the Internet of Things, power big data is growing rapidly. By mining power big data, we can not only meet the needs of power industry development, but also serve the needs of economic development.

[0003] Artificial intelligence is to make computers imitate human logical thinking and advanced wisdom, which can be divided into three levels: computational intelligence, perceptual intelligence and cognitive intelligence. Computational intelligence is to enable machines / computers to have high-performance computing capabilities, even surpassing human computing capabilities to process massive data; perceptual intelligence is to enable machines to perceive the surrounding environment like humans, including hearing, vision, touch, etc.; cognitive intelligence is to enable machines to have human rational thinking ability and make correct decisions. The integration of the three capabilities ultimately allows machines to achieve human-like intelligence to fully assist or even replace human work. With the help of artificial intelligence technology, the photovoltaic power generation and load demand of each user in the community microgrid can be predicted.

[0004] Mathematical modeling is to establish mathematical models based on practical problems, solve mathematical models, and then solve practical problems based on the results. The modeling process includes model preparation, model assumptions, model establishment, model solving, and model verification. Model preparation is to further describe the problem in mathematical language based on the understanding of the actual background of the problem; model assumptions are to make some appropriate assumptions based on the purpose of modeling and simplify the problem as necessary; model establishment is to use appropriate mathematical tools to characterize the mathematical relationship between various variables and constants and establish the corresponding mathematical structure; model solving is to use some algorithms to calculate all parameters in the model; model verification is to compare and analyze the model calculation results with the actual results to verify the accuracy of the model.

[0005] The community microgrid architecture currently used in the power trading system usually requires the installation of energy storage devices on the user side. However, the cost of energy storage devices is relatively high. Without strong government subsidies to promote installation, it is difficult to attract users to install energy storage devices on their own. Therefore, a novel energy storage architecture is needed. In addition, the current trading mechanisms in the power trading system are mainly divided into centralized and point-to-point. The point-to-point mechanism is very complex and difficult to implement in small-scale markets such as community microgrids. The existing centralized mechanism lacks fair pricing strategies and lacks protection for user privacy. Therefore, an improved trading mechanism is needed to overcome the above shortcomings. Finally, the existing power trading system is based on assumptions and does not consider prediction errors. Therefore, a model for handling power prediction errors is needed. Summary of the invention

[0006] In view of the shortcomings of the prior art described above, an object of the present invention is to provide a power supply and demand balancing system, method and application thereof, which are used to solve the technical problem of unbalanced power supply and demand for each user in a microgrid in a power trading system in the prior art.

[0007] To achieve the above objectives and other related objectives, the present invention provides a method for balancing power supply and demand, the method comprising:

[0008] Predicting the power generation and / or load demand of each user in the microgrid in the next transaction time window according to the preset time window of the microgrid power supply and demand balancing system, and dividing all users in the microgrid into a power seller set and a power buyer set based on the obtained power generation and / or load demand of each user;

[0009] Establishing a power purchase cost model and a seller benefit model, and calculating the power purchase cost of the system dispatcher and the benefit of the seller through the power purchase cost model and the seller benefit model;

[0010] Determining the electricity collection value collected from each electricity seller under the conditions of minimizing the electricity purchase cost of the system dispatcher and maximizing the benefit of the seller;

[0011] Establishing a buyer benefit model, and calculating the total buyer benefit through the buyer benefit model;

[0012] Determining the power allocation value allocated to each power buyer on the condition of maximizing the sum of the benefits of the buyers;

[0013] Electricity is collected according to the electricity collection value determined by each electricity seller, and electricity is distributed according to the electricity distribution value allocated by each electricity buyer, so as to achieve the balance of power supply and demand in the microgrid.

[0014] In one embodiment, the step of respectively predicting the power generation and load demand of each user in the microgrid in the next transaction time window according to the preset time window of the system includes:

[0015] The power generation and load demand of each user in the microgrid are collected in real time as power big data;

[0016] Preprocessing the power big data to select appropriate features and create training samples and test samples;

[0017] Training a user power prediction model based on the training samples;

[0018] Evaluating the user power prediction model through the test sample to obtain an optimal user power prediction model;

[0019] The optimal user power forecasting model is used to predict the user's power generation and / or load demand in the next transaction time window.

[0020] In one embodiment, the method further includes performing compensation settlement on the trading subject according to the prediction error.

[0021] In one embodiment, the step of performing compensation settlement on the trading subject according to the prediction error includes:

[0022] Compensation settlement is carried out based on the difference between the actual power sold by each of the power sellers and the power collected from the sellers; and / or

[0023] Compensation settlement is performed based on the difference between the actual purchased power of each of the power buyers and the power allocated to the buyer.

[0024] In one embodiment, the method further comprises adaptively adjusting the transaction time window according to the prediction error and the system operation pressure.

[0025] In order to achieve the above-mentioned object and other related objects, the present invention further provides a power supply and demand balancing system, the power supply and demand balancing system comprising:

[0026] The power prediction module is used to predict the power generation and / or load demand of each user in the microgrid in the next trading time window according to the preset time window of the microgrid trading system, and divide all users in the microgrid into a power seller set and a power buyer set based on the obtained power generation and / or load demand of each user;

[0027] an electricity collection module, used to establish an electricity purchase cost model and a seller benefit model, and calculate the electricity purchase cost of the system dispatcher and the benefit of the seller through the electricity purchase cost model and the seller benefit model, determine the electricity collection value collected from each electricity seller under the condition of minimizing the electricity purchase cost of the system dispatcher and maximizing the benefit of the seller, and perform electricity collection according to the determined electricity collection value collected by each electricity seller; and

[0028] The power distribution module is used to establish a buyer benefit model and calculate the sum of the buyer's benefits through the buyer benefit model, determine the power distribution value allocated to each power buyer on the condition of maximizing the sum of the buyer's benefits, and distribute power according to the power distribution value allocated to each power buyer to achieve a balance between power supply and demand in the microgrid.

[0029] In one embodiment, the power prediction module includes:

[0030] A data loading submodule is used to obtain the power generation and load demand of each user in the microgrid in real time as power big data;

[0031] A preprocessing submodule, used for preprocessing the power big data to select appropriate features and create training samples and test samples;

[0032] A training submodule, used for training a user power prediction model based on the training samples;

[0033] An evaluation submodule, used for evaluating the user power prediction model through the test sample to obtain an optimal user power prediction model;

[0034] The prediction submodule is used to predict the power generation and / or load demand of the user in the next transaction time window through the optimal user power prediction model.

[0035] In one embodiment, the system further includes a prediction error processing module for performing compensation settlement on the transaction subject according to the prediction error.

[0036] In one embodiment, the prediction error processing module includes:

[0037] for performing compensation settlement according to the difference between the actual power sold by each of the power sellers and the power collected from the power sellers; and / or

[0038] A buyer compensation settlement submodule is used to perform compensation settlement according to the difference between the actual purchased power of each power buyer and the power allocated to the power buyer.

[0039] In one embodiment, the microgrid trading system further includes a time window adjustment module, which is used to adjust the time window according to the prediction error and the system operation pressure.

[0040] In order to achieve the above-mentioned object and other related objects, the present invention also provides a microgrid, the microgrid comprising

[0041] A system dispatcher connected to the main power grid, the system dispatcher comprising a cloud energy storage device and any one of the above-mentioned power supply and demand balancing systems;

[0042] The user side is connected to the system dispatcher and includes a number of users, each of which is equipped with an electronic terminal, and a load and a power generation device connected to the system dispatcher through the electronic terminal;

[0043] The system dispatcher performs power trading by controlling the power supply and demand balancing system.

[0044] The present invention solves the problem of unbalanced power supply and demand for a single user by establishing a reliable and stable power supply and demand balancing system in a community microgrid;

[0045] Compared with the current method of connecting surplus power to the grid, it not only greatly increases the seller's profit from electricity sales, but also greatly reduces the buyer's electricity purchase cost and the interference to the main power grid;

[0046] The present invention proposes a compensation and settlement model for power forecast errors, which solves the problem of economic losses to trading entities caused by forecast errors. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 Shown is a structural block diagram of the microgrid of the present invention.

[0048] Figure 2 Shown is a structural block diagram of the power trading system of the present invention.

[0049] Figure 3 Shown is a schematic diagram of the predictive transaction process based on the time window of the present invention.

[0050] Figure 4 Shown is a schematic flow chart of the power trading method of the present invention.

[0051] Figure 5 Shown is a schematic diagram of a prediction model learning algorithm of the power prediction module of the present invention.

[0052] Figure 6 Shown is a schematic diagram of electricity trading between users in the community microgrid of the present invention.

[0053] Figure 7 It is a schematic diagram showing the relationship between the convergence time of the optimal electricity pricing algorithm of the present invention and the total number of users in the microgrid.

[0054] Figure 8It is a schematic diagram showing the relationship between the convergence time of the power distribution algorithm of the present invention and the total number of users in the microgrid.

[0055] Fig. 9 Shown is a compensation settlement flow chart for the electricity seller of the present invention.

[0056] Fig.10 Shown is a compensation settlement flow chart for the electricity buyer of the present invention.

[0057] Fig.11 Shown is a structural block diagram of the service device of the present invention.

[0058] Fig.12 Displayed is user power generation prediction data in an example of the present invention.

[0059] Fig.13 Displayed is the user's electricity consumption forecast data in the example of the present invention.

[0060] Fig.14 It shows a schematic diagram comparing the impact of the power trading proposed by the present invention and the surplus power access mode on the stability of the main power grid.

[0061] Fig.15 It shows a schematic diagram comparing the buyer's electricity purchase cost under the electricity transaction proposed by the present invention and the surplus power grid-connected mode.

[0062] Fig.16 It shows a schematic diagram comparing the seller's electricity sales profit under the electricity transaction proposed by the present invention and the surplus power grid-connected mode. DETAILED DESCRIPTION

[0063] The following describes the embodiments of the present invention through specific examples, and those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention.

[0064] See also Figure 1-16 It should be noted that the illustrations provided in this embodiment are only used to illustrate the basic concept of the present invention in a schematic manner, and the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0065] Figure 1 Figure 2 shows a structural block diagram of a community microgrid. Figure 1As shown, the microgrid includes a system dispatcher 200 and a user side 300 that are interconnected; the system dispatcher 200 is connected to the main grid 400, and can purchase electricity from the main grid 400 or sell excess electricity to the main grid 400; the system dispatcher 200 is used to complete electricity trading by dispatching electricity; the user side 300 includes a number of community microgrid users, each of which is equipped with an electronic terminal 5, and a load 6 and a power generation device 7 connected to the system dispatcher 200 through the electronic terminal 5, the load 6 refers to the user's power consumption device, the power generation device 7 is a power production device, the power generation device 7 can be, for example, a photovoltaic power generation device or a wind power generation device, the electronic terminal 5 is used to measure and record the power data produced and consumed by the user in real time, and when necessary, transmit the recorded data to the system dispatcher 200, in the present invention, the electronic terminal 5 can be, for example, a bidirectional metering smart meter.

[0066] like Figure 1 and Figure 2 As shown, in order to solve the problem of unbalanced power supply and demand of a single user in the process of power trading, an embodiment of the present invention discloses a centralized power trading system 100 based on big data and artificial intelligence in a community microgrid. The power trading system 100 can be built into the system dispatcher 200, and is centrally controlled by the system dispatcher 200 to complete power trading. The power trading system 100 mainly includes three progressive modules, namely, a power prediction module 1, a power trading module 2 and a prediction error processing module 3. Among them, the power prediction module 1 is used to predict the power generation and load demand of each user in the microgrid in the next trading time window according to the preset time window of the microgrid trading system, and divide all users in the microgrid into a power seller set and a power buyer set based on the power generation and load demand of each user obtained; the power trading module 2 includes a power collection module and a power distribution module, the power collection module is used to establish a power purchase cost model and a seller benefit model, and calculate the power purchase cost and seller benefit of the system dispatcher 200 through the power purchase cost model and the seller benefit model, and collect power from each power seller on the condition of minimizing the power purchase cost of the system dispatcher 200 and maximizing the benefit of the seller, the power distribution module is used to establish a buyer benefit model, and calculate the sum of the buyer's benefits through the buyer benefit model, and allocate power to the power buyer on the condition of maximizing the sum of the buyer's benefits; the prediction error processing module 3 is used to compensate and settle the transaction subject according to the prediction error. The specific operations performed by the power prediction module 1, the power trading module 2 and the prediction error processing module 3 will be described in detail below. Figure 4 Provide detailed explanation.

[0067] It should be noted that the power trading system 100 and method of the present invention are based on a cloud energy storage architecture and a time window prediction trading model. The cloud energy storage architecture means that no energy storage device is installed on the user side 300, but a larger cloud energy storage device 8 is installed on the system scheduler 200 side, and each user shares the cloud energy storage device 8. Figure 3 A schematic diagram of the time window prediction transaction process is shown in FIG. Figure 3 As shown, each trading time period is divided into three stages, namely, the trading mechanism operation stage (first stage), the contract execution stage (second stage) and the prediction error processing stage (third stage). In the time period [Δt, 2Δt][Δt, 2Δt], the trading mechanism operation stage determines how the seller and the buyer trade by predicting the power supply and demand of each user in the community microgrid in the next time period [2Δt, 3Δt][2Δt, 3Δt], and then the two parties sign a power trading contract. Then, in the contract execution stage in the time period [2Δt, 3Δt][2Δt, 3Δt][2Δt, 3Δt, [2Δt, 3Δt], the system scheduler 200 completes the power transaction between users by dispatching power. Finally, in the prediction error processing stage in [2Δt, 3Δt][2Δt, 3Δt], the system processes the power prediction error of the previous trading time period (that is, the time period [Δt, 2Δt][Δt, 2Δt]).

[0068] Figure 4 Is Figure 1 The power trading system 100 in the present invention performs a flowchart of the power trading method of the present invention. Figure 4 The power trading method is described.

[0069] In step S1, the system is initialized to initialize the time window Δt, the error threshold δ, and the system pressure threshold ε. The initial value of the error Err is set to 0, and the initial value of the system operating pressure P is set to 0.

[0070] In step S2, it is determined whether the error value Err and the system operating pressure P are simultaneously smaller than their respective threshold values ​​δ and ε.

[0071] If the determination result is "yes", execute step S3.

[0072] If the determination result is "No", execute step S8.

[0073] In step S3, power forecasting is performed, and the power generation and load demand of each user in the microgrid in the next trading time window are predicted according to the preset time window of the microgrid power trading system 100 (step S31), and based on the obtained power generation and load demand of each user, all users in the microgrid are divided into a set of power sellers and a set of power buyers (step S32).

[0074] In step S31, the steps of respectively predicting the power generation and load demand of each user in the microgrid in the next trading time window according to the preset time window of the microgrid power trading system 100 include:

[0075] Step S311: Use the set N to represent all users in the community microgrid:

[0076] N={1,2,…,i,…,n}

[0077] Where i is the user number in the community microgrid (the same reference is used unless otherwise specified in the following text);

[0078] Step S312: The system collects the power data produced and consumed by users measured by the two-way metering smart meter in real time, and forms a minute-level historical record of each user's power generation and power consumption as power big data;

[0079] Step S313: the power prediction module 1 pre-processes the power big data recorded in the previous step, selects appropriate features and creates a large-scale sample. The pre-processing process may include clustering, pre-filtering, normalization and feature selection steps;

[0080] Step S314: the power prediction module 1 uses a radial basis function neural network to train a user power prediction model based on the training samples. Specifically, the power prediction module 1 uses a radial basis function neural network to train user power generation and power consumption prediction models based on the training samples.

[0081] Step S315: the power prediction module 1 evaluates the model through the test samples to obtain the optimal user power prediction model. Specifically, the power prediction module 1 evaluates the model through the test samples to obtain the optimal photovoltaic power generation prediction model and load demand prediction model;

[0082] Step S316: The photovoltaic power generation prediction model predicts that the photovoltaic power generation of user i in the next transaction time window t is The load demand forecasting model predicts that the load demand of user i in the next transaction time window t is

[0083] Figure 5 FIG. 2 shows a schematic diagram of the prediction model learning algorithm of the power prediction module 1. Figure 5As shown, the power prediction module 1 includes a power data loading submodule 11, which is used to obtain the power generation and load demand of each user in the microgrid in real time as power big data (corresponding to step S312); a preprocessing submodule 12, which is used to preprocess the power big data to select appropriate features and create training samples and test samples (corresponding to step S313); a training submodule 13, which is used to train the user power prediction model based on the training samples (corresponding to step S314); an evaluation submodule 14, which is used to evaluate the power prediction model through the test samples to obtain the optimal user power prediction model (corresponding to step S315); a prediction submodule 15, which is used to predict the power generation and load demand of the user in the next trading time window through the optimal power prediction model (corresponding to step S316). The user power prediction model includes the above-mentioned user power generation prediction model and the above-mentioned user power consumption prediction model, and the optimal user power prediction model includes the above-mentioned photovoltaic power generation prediction model and the above-mentioned load demand prediction model.

[0084] In step S32, based on the acquired power generation and load demand of each user, the step of dividing all users in the microgrid into a power seller set and a power buyer set includes:

[0085] Step 321: Based on the prediction in step S31 and Calculate the net output power of user i in the next transaction time window t

[0086] Step 322: When When user i is the electricity buyer, When user i is the electricity seller, When , the power supply and demand of user i is balanced, so it does not participate in power trading;

[0087] Step 323: All users in the community microgrid are divided into a power buyer set BB and a power seller set SS, where

[0088] In step S4, power trading is carried out through the power trading module 2. The present invention proposes a centralized power trading scheme. The power trading is divided into two stages. First, the system scheduler 200 collects power from the seller set (corresponding to step S41 below), and then the system scheduler 200 distributes power to the buyer set (corresponding to step S42 below). That is, the power trading module 2 includes a power collection module and a power distribution module. Among them, the power trading diagram between users in the community microgrid is as follows Figure 6 shown.

[0089] In step S41, during the electricity collection stage, the electricity collection module establishes an electricity purchase cost model and a seller benefit model, and calculates the electricity purchase cost of the system scheduler 200 and the seller's benefit through the electricity purchase cost model and the seller benefit model, and collects electricity from each electricity seller on the condition of minimizing the electricity purchase cost of the system scheduler 200 and maximizing the seller's benefit.

[0090] Specifically, step S41 includes the following sub-steps:

[0091] Step S411: Establishing a power purchase cost model of the system scheduler 200:

[0092]

[0093]

[0094] Where: pp is the internal power transaction price proposed by the system dispatcher 200, and the power purchase cost C of the system dispatcher 200 DSO C DSO It includes two parts. The item on the left of the plus sign represents the cost of the system dispatcher 200 purchasing electricity from all sellers in the microgrid, and the item on the right represents the cost of the system dispatcher 200 purchasing additional required electricity from the main grid 400;

[0095] Step S412: Establishing a benefit model for the electricity seller:

[0096] U i =k i ln(1+D i )+p(G i -D i )

[0097] Constraints:

[0098] The benefit of seller i is U i U i It consists of two parts. The item on the left of the plus sign indicates that seller i consumes electricity D i D i The satisfaction gained, i k i is the electricity efficiency parameter of seller i, the term on the right of the plus sign is the profit obtained by seller i by selling excess electricity to the system dispatcher 200, and is the minimum load requirement that seller i must meet;

[0099] Step S413: Mathematically formally define the optimization problem of the power collection phase of the system scheduler 200:

[0100] Objective function:

[0101]

[0102]

[0103] Constraints:

[0104] The system dispatcher 200 seeks an optimal internal power transaction price with the goal of minimizing the power purchase cost and maximizing the seller's benefit;

[0105] Step S414: using a distributed iterative algorithm to solve the optimization problem in the power collection phase:

[0106] The input of the algorithm includes the main grid 400 electricity sales price and electricity purchase price The seller set S and the buyer set B, and their corresponding photovoltaic power generation G i G i and load demand D i D i , the seller's electricity efficiency parameter k i k i , the output is the optimal electricity transaction price p within the community microgrid * p * , the specific process is as follows:

[0107]

[0108] First, the initialization work is performed, the optimal internal power transaction price is set as the power purchase price of the main grid 400, and the minimum power purchase cost of the system dispatcher 200 is set as the price of purchasing all the required power from the main grid 400 (Line 1-2), and then the cyclic iteration process is started. Within the range, traverse pp so that Minimum (Line 3 to 14), the p returned by the algorithm at the end * p * is the optimal internal electricity trading price. Find the optimal internal electricity trading price p * p * After that, the energy consumed by seller i in response to the optimal electricity price to maximize the profit of electricity sales can be obtained. Then the system scheduler 200 collects of electrical energy. Figure 7 The figure shows the relationship between the convergence time of the optimal electricity pricing algorithm (distributed iterative algorithm) and the number of users in the community microgrid. Figure 7 It can be seen that the convergence time of the optimal electricity pricing algorithm is proportional to the number of users in the community microgrid.

[0109] In step S42, during the power distribution phase, a buyer benefit model is established through the power distribution module, and the sum of the buyer's benefits is calculated through the buyer benefit model, so that power is distributed to the power buyer on the condition of maximizing the sum of the buyer's benefits.

[0110] Specifically, the step S42 includes the following sub-steps:

[0111] Step S421: Establishing a comprehensive scoring model for electricity buyers:

[0112]

[0113] The comprehensive score PI of buyer i i PI i It consists of three parts, including the historical electricity contribution ratio of buyer i Current electricity demand share and reliability C i C i is the total amount of electricity sold by buyer i to the community microgrid during the 200 statistical cycles of the system dispatcher, C Total C Total is the total amount of electricity sold by all users to the community microgrid within the 200 statistical cycles of the dispatcher, D i D i is the amount of electricity that buyer i needs to purchase during the current trading period, D Total D Total is the total amount of electricity that all users need to purchase during the current trading period t, N f N f N represents the number of times buyer i successfully executes the contract during the statistical period. Total N Total represents the total number of times the buyer i executes the contract in the statistical period, λ(λ>0),μ(μ>0)λ(λ>0),μ(μ>0) are weight factors, which are set by the system scheduler 200, and represent the importance of contribution, local load demand and reliability respectively;

[0114] Step S422: Establishing a benefit model for the electricity buyer based on the comprehensive scoring model:

[0115]

[0116] Among them, F i F i is the benefit to buyer i, PI i PI i is the comprehensive score of buyer i, s i s i is the amount of electricity requested by buyer i from the system scheduler 200 according to its own strategy, EAi EA i is the amount of electricity allocated to buyer i by the system scheduler 200 according to a certain power allocation strategy;

[0117] Step S423: Mathematically define the optimization problem of the power allocation phase of the system scheduler 200:

[0118] Objective function:

[0119]

[0120] Constraint: 0≤EA i ≤s i 0≤EA i ≤s i

[0121]

[0122] Buyer i's optimal power request strategy, is the optimal amount of electricity allocated to buyer i by the system scheduler 200,

[0123] The goal of the system in the power distribution phase is to maximize social welfare, that is, the sum of the benefits of all power buyers. Constraint 1 represents the power EA allocated to the buyer by the system scheduler 200. i No more than the electricity requested by the buyer i ,Constraint 2 indicates that the total amount of power allocated to all buyers by the system scheduler 200 cannot exceed the total amount of power purchased from the seller set during the power collection phase

[0124] Step S424: The water injection algorithm solves the optimization problem of the power distribution stage:

[0125]

[0126]

[0127] The algorithm first performs initialization (Line 1-4), then starts a loop iteration process until the remaining power to be allocated is 0 (Line 5-17), and finally returns the water volume set EAEA allocated to all water tanks, that is, the power set EAEA allocated to all buyers (Line 18), that is, the set of optimal power allocated to each buyer by the system scheduler 200. The system scheduler 200 allocates electric energy to buyers according to the optimal power corresponding to each buyer in the power set EAEA. Figure 8 The figure shows the relationship between the convergence time of the optimal power allocation algorithm (water injection algorithm) and the number of users in the community microgrid. Figure 8It can be seen that the convergence time of the optimal power allocation algorithm is proportional to the number of users in the community microgrid.

[0128] In step S5, the prediction error processing module 3 performs error processing to compensate and settle the transaction subject according to the prediction error. Specifically, the prediction error processing module 3 solves the problem of economic losses of the transaction subject caused by the power prediction error by running the compensation settlement model. The compensation settlement model consists of the seller compensation settlement model in the power collection stage and the buyer compensation settlement model in the power distribution stage. Fig. 9 A flowchart of executing the seller compensation liquidation model for compensation liquidation is shown (corresponding to step S51 below). Fig.10 It is a flow chart showing the execution of the buyer's compensation settlement model for compensation settlement (corresponding to step S52 below).

[0129] like Fig. 9 As shown, in step S51, the following steps are included:

[0130] Step S511, seller j (j∈S) signs a contract with DSO (system scheduler 200) based on the forecast;

[0131] Step S512, seller j (j∈S) and DSO execute the contract based on the actual power state;

[0132] Step S513: After the two parties have completed the contract execution in the power collection phase, the actual photovoltaic power generation G of the seller j in the transaction time period t is obtained by reading the two-way metering smart meter. j,act G j,act and load demand D j,act D j,act ;

[0133] Step S514, calculating the error ΔE j,sell ΔE j,sell , according to the actual photovoltaic power generation G of seller j in the transaction time period t j,act G j,act and load demand D j,act D j,act , we can find out the actual electricity sold by seller j It is expressed as:

[0134]

[0135] Therefore, the difference between the actual power sold by seller j and the predicted power sold is ΔE j,sell ΔE j,sell Expressed as

[0136] Step 515: When ΔE j,sell =0ΔEj,sell = 0, the seller j just provides the system dispatcher 200 with the predicted sellable electricity during the trading period t, so the settlement is normal. j,sell >0ΔE j,sell > 0, the seller j not only provides the dispatcher 200 with the predicted sellable power in full during the trading period t, but also provides an additional ΔE j,sell ΔE j,sell Therefore, the scheduler 200 needs to compensate the profit of seller j, and the profit of seller j will increase by ΔE j,sell *p gb ΔE j,sell *p gb , when ΔE j,sell <0ΔE j,sell When <0, the seller j does not provide the dispatcher 200 with the predicted sellable electricity in the trading period t, and provides less ΔE j,sell ΔE j,sell Under the assumption that other users and the dispatcher 200 conduct error-free power transactions based on the prediction, the dispatcher 200 will purchase the missing power ΔE from the main grid 400 to maintain the power supply and demand balance within the community microgrid. j,sell *p gs ΔE j,sell ΔE j,sell ΔE j,sell *p gs ΔE j,sell ΔE j,sell , so seller j needs to compensate the loss of scheduler 200, and the profit of seller j will be reduced by ΔE j,sell *p gs ΔE j,sell *p gs ;

[0137] Step S516, settlement results.

[0138] like Fig.10 As shown, in step S52, the following steps are included:

[0139] Step S521, buyer i (i∈S) signs a contract with DSO (system scheduler 200) based on the forecast;

[0140] Step S522, seller i (i∈S) and DSO execute the contract based on the actual power state;

[0141] Step S523: After the two parties have completed the execution of the contract in the power distribution phase, the actual photovoltaic power generation G of buyer i in the transaction time period t is obtained by reading the two-way metering smart meter. i,act G i,act and load demand D i,act Di,act ;

[0142] Step S524, calculate the error ΔE i,req ΔE i,req Specifically, the actual photovoltaic power generation G of buyer i during the transaction period t is i,act G i,act And load demand, we can find the actual power purchased by buyer i It is expressed as:

[0143]

[0144] Therefore, the difference between the actual power purchased by buyer i and the predicted power purchased is ΔE i,req ΔE i,req Expressed as

[0145] Step S525, when ΔE i,req =0ΔE i,req = 0, it is obvious that buyer i has just purchased the predicted power demand from the system scheduler 200 during the transaction period t, so the settlement is normal. i,req >0ΔE i,req > 0, buyer i not only purchases the predicted power demand from the system scheduler 200 during the trading period t, but also purchases ΔE i,req ΔE i,req At this time, the system dispatcher 200 needs to purchase ΔE from the main grid 400. i,req ΔE i,req Therefore, buyer i needs to compensate the system dispatcher 200 for the power purchase cost, that is, the power purchase cost of buyer i will increase by ΔE i,req *p gs ΔE i,req *p gs , when ΔE i,req <0ΔE i,req When ΔE is less than 0, buyer i does not purchase the predicted power from the system scheduler 200 in full during the transaction period t, and purchases less than ΔE i,req ΔE i,req Under the assumption that other users and the system dispatcher 200 conduct error-free power transactions based on the prediction, the system dispatcher 200 will sell power ΔE to the main grid 400 to maintain the power supply and demand balance within the community microgrid. i,req ΔE i,req Therefore, the revenue obtained by the system dispatcher 200 from selling electricity should be compensated to buyer i, that is, the electricity purchase cost of buyer i will be reduced by ΔE i,req *p gb ΔE i,req *p gb .

[0146] Step S6, determining whether the system termination command is satisfied.

[0147] If the determination result is "yes", the step ends.

[0148] If the determination result is "No", it is necessary to go through step S7.

[0149] In step S7, the error Err and the system operating pressure P in step 2 are updated. Specifically, the system operating pressure is monitored, and the error Err and the system operating pressure P in step 2 are updated using the system operating pressure obtained by monitoring and the error obtained by calculation, and step S2 is re-executed. If the judgment result of step S2 is "no", step S8 needs to be executed.

[0150] In step S8, the transaction time window Δt is adjusted (for example, reduced by a preset value based on the original transaction time window) and the process returns to step S3 to predict the photovoltaic power generation and load demand of each residential user in the community microgrid in the next transaction time window based on the adjusted transaction time window Δt, and execute subsequent steps.

[0151] In order to balance the prediction error and the system pressure, by adding the time window adjustment module 4 to the power trading system 100 (by adding steps S7, S2 and S8), the trading time window can be adjusted by using the system operation pressure obtained by monitoring and the error obtained by calculation until it is adjusted to the optimal trading time window. That is, the power trading system 100 of the present invention has the characteristics of dynamic self-adaptation. In other words, the power trading system 100 of the present invention operates in a closed loop and can be adjusted to the optimal trading time window.

[0152] It should be noted that if Fig.11 As shown, the power trading method and system of this embodiment can be implemented by a service device 1, which includes a memory 503, a processor 501 and a communicator 502 that are interconnected. The memory 501 stores a computer program, which implements the web page text processing method when executed by the processor 501.

[0153] The above-mentioned processor 501 can be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it can also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field programmable gate array (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components; the above-mentioned memory 503 may include random access memory (RAM), and may also include non-volatile memory, such as at least one disk storage.

[0154] It should be noted that the computer program in the above-mentioned memory 503 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when it is sold or used as an independent product. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or the part of the technical solution can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for a computer device (which can be a personal computer, an electronic device, or a network device, etc.) to perform all or part of the steps of the methods of various embodiments of the present invention.

[0155] In an embodiment of the present invention, a computer storage medium may also be provided, which stores a computer program, and when the program is executed by a processor, the weld material quantity control method is implemented; the computer storage medium includes all forms of non-volatile memory, media and memory devices, including, for example: semiconductor memory devices, such as EPROM, EEPROM and flash memory devices; magnetic disks, such as internal hard disks or removable disks; magneto-optical disks; and CD-ROM and DVD-ROM disks.

[0156] The present invention will be described below with reference to a specific example.

[0157] System initialization, consider a community microgrid consisting of 6 residential users, each of which is equipped with a photovoltaic power generation system and a bidirectional smart meter. The main grid's electricity price of 400 p gs p gs Assumed to be 0.92 yuan / kWh, the power purchase price of the main grid 400 p gb p gbThe initial transaction time window set by the system scheduler 200 is 0.37 yuan / kWh, and the initial transaction time window is 1 hour. The power prediction module 1 predicts the photovoltaic power generation and load demand of each residential user in the community microgrid (i.e., the user dynamic model) based on power big data and artificial intelligence technology. The schematic diagram of the prediction model learning algorithm is shown in Figure 5 As shown. Using the prediction model, the power generation G predicted by each user i and power consumption D i Respectively Fig.11 and Fig.12 shown.

[0158] A prediction-based user dynamic model is obtained, and all residential users in the community microgrid are divided into a set of electricity sellers and a set of electricity buyers.

[0159] The power trading module 2 solves the problem of trading power between the seller set and the buyer set, which is divided into a stage where the system scheduler 200 collects power from the seller set and a stage where the system scheduler 200 distributes power to the buyer set. The power trading diagram is shown in FIG. Figure 6 shown.

[0160] In the power collection stage, by establishing the power purchase cost model of the system scheduler 200 and the benefit model of the seller, the system collects power with the goal of minimizing the power purchase cost of the system scheduler 200 and maximizing the benefit of the seller. The optimal power pricing problem in the power collection stage is formally defined as:

[0161] Objective function:

[0162]

[0163]

[0164] Constraints:

[0165] Then, the optimal electricity transaction price p* of this problem is solved by the distributed iterative algorithm proposed above in the present invention. The relationship between the convergence time of the optimal electricity pricing algorithm and the number of users in the community microgrid is shown in the figure below: Figure 7 shown.

[0166] Table 1 shows the power collection process in the 11th trading period:

[0167]

[0168] Among them, is p * is the optimal electricity price obtained by solving the optimal electricity pricing problem, is the electric energy consumed by seller i in response to the optimal electricity price in order to maximize the profit of electricity sales. It is the electric energy collected by the system scheduler 200 from seller i.

[0169] Step 6: In the power distribution stage, by establishing the buyer's benefit model, the system distributes power with the goal of maximizing social welfare (the sum of the benefits of all buyers). The optimal power distribution problem in the power distribution stage is formally defined as:

[0170] Objective function:

[0171]

[0172] Constraint: 0≤EA i ≤s i 0≤EA i ≤s i

[0173]

[0174] Then, the optimal power allocation strategy for this problem is solved by the water injection algorithm proposed in the present invention. The relationship between the convergence time of the optimal power allocation algorithm and the number of users in the community microgrid is shown in the figure below: Figure 8 shown.

[0175] Table 2 shows the process of allocating electricity in the 11th trading period:

[0176]

[0177] in, Buyer i's optimal power request strategy, is the optimal power allocated to buyer i by the system scheduler 200, E ex It is the difference between the total amount of electricity purchased from the set of sellers and the total amount of electricity allocated to buyers. For other definitions of compliance, please refer to the relevant parts above and will not be elaborated here.

[0178] The prediction error processing module 3 runs a compensation settlement model to solve the problem of economic losses caused to the trading subject by the power prediction error. The compensation settlement model is composed of the seller's compensation settlement model in the power collection stage (such as Fig. 9 The compensation and liquidation model of the buyer in the power distribution stage consists of Fig.10 shown).

[0179] The power trading system 100 continuously monitors the system operating pressure and calculates the power forecast error to find the trading time window when the operating pressure and the forecast error are optimally balanced, so as to balance the system operating pressure and the forecast error.

[0180] The present invention compares the surplus power access to the grid from three aspects: the stability of the main grid 400, the buyer's power purchase cost and the seller's power sales profit. Under the power trading mode, the power exchange between the microgrid and the main grid 400 is reduced by 72.6% on average, the buyer's power purchase cost is reduced by 46.2% on average, and the seller's power sales profit is increased by 62.5% on average, as shown in the figure below. Fig.11 , Fig.12 and Fig.13 shown.

[0181] The present invention proposes a centralized power trading system 100 based on big data and artificial intelligence in a community microgrid, which solves the problem of unbalanced power supply and demand for a single user in a community microgrid. The system has been proven to be stable and reliable through experiments, and the power pricing algorithm and allocation algorithm adopted are simple and efficient. Compared with the currently adopted surplus power grid connection, it not only greatly increases the seller's power sales profit, but also greatly reduces the buyer's power purchase cost and interference to the main power grid 400.

[0182] In the description herein, many specific details, such as examples of components and / or methods, are provided to provide a complete understanding of embodiments of the present invention. However, those skilled in the art will recognize that embodiments of the present invention may be practiced without one or more of the specific details or with other devices, systems, components, methods, components, materials, parts, etc. In other cases, well-known structures, materials, or operations are not specifically shown or described in detail to avoid obscuring aspects of embodiments of the present invention.

[0183] References throughout the specification to "one embodiment," "an embodiment," or "a specific embodiment" mean that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the invention, and not necessarily in all embodiments. Thus, various appearances of the phrases "in one embodiment," "in an embodiment," or "in a specific embodiment" in different places throughout the specification do not necessarily refer to the same embodiment. In addition, the particular features, structures, or characteristics of any specific embodiment of the invention may be combined with one or more other embodiments in any suitable manner. It should be understood that other variations and modifications of the embodiments of the invention described and illustrated herein may be possible in light of the teachings herein and are to be considered part of the spirit and scope of the invention.

[0184] It should also be understood that one or more of the elements shown in the figures may also be implemented in a more separate or more integrated manner, or even removed because they are inoperable in certain circumstances or provided because they may be useful depending on the application.

[0185] In addition, unless otherwise explicitly stated, any marking arrows in the drawings should be regarded as exemplary only and not limiting. In addition, unless otherwise specified, the term "or" used herein is generally intended to mean "and / or". In the case where the term is unclear because it is anticipated that the ability to separate or combine is provided, the combination of components or steps will also be regarded as specified.

[0186] As used in the description herein and throughout the claims that follow, “a,” “an,” and “the” include plural references unless otherwise indicated. Likewise, as used in the description herein and throughout the claims that follow, the meaning of “in” includes “in” and “on,” unless otherwise indicated.

[0187] The above description of the illustrated embodiments of the present invention (including the contents described in the Abstract) is not intended to be an exhaustive list or to limit the present invention to the precise form disclosed herein. Although specific embodiments of the present invention and examples of the present invention are described herein for illustrative purposes only, various equivalent modifications are possible within the spirit and scope of the present invention as will be recognized and appreciated by those skilled in the art. As noted, these modifications may be made to the present invention in accordance with the above description of the embodiments described in the present invention, and these modifications will be within the spirit and scope of the present invention.

[0188] Systems and methods have been generally described herein as details that aid in understanding the present invention. In addition, various specific details have been given to provide an overall understanding of embodiments of the present invention. However, those skilled in the relevant art will recognize that embodiments of the present invention may be practiced without one or more of the specific details, or may be practiced using other devices, systems, accessories, methods, components, materials, parts, etc. In other cases, well-known structures, materials, and / or operations are not specifically shown or described in detail to avoid confusion with various aspects of embodiments of the present invention.

[0189] Thus, although the invention has been described herein with reference to specific embodiments thereof, freedom of modification, various changes and substitutions are intended within the foregoing disclosure, and it should be understood that in some cases, some features of the invention will be employed without the corresponding use of other features without departing from the scope and spirit of the proposed invention. Thus, many modifications may be made to adapt a particular environment or material to the essential scope and spirit of the invention. The invention is not intended to be limited to the specific terms used in the claims below and / or the specific embodiments disclosed as the best mode contemplated for carrying out the invention, but the invention will include any and all embodiments and equivalents falling within the scope of the appended claims. Thus, the scope of the invention will be determined solely by the appended claims.

Claims

1. A microgrid, It is characterized in that include: A system dispatcher connected to the main power grid, the system dispatcher comprising a cloud energy storage device and a power supply and demand balancing system; The power supply and demand balancing system comprises: The power prediction module is used to predict the power generation and / or load demand of each user in the microgrid in the next trading time window according to the preset time window of the microgrid trading system, and divide all users in the microgrid into a power seller set and / or a power buyer set based on the obtained power generation and load demand of each user; The method of predicting the power generation and / or load demand of each user in the microgrid in the next trading time window according to the preset time window of the microgrid trading system includes: The system collects the power data produced and consumed by users measured by the two-way metering smart meter in real time, and forms a minute-level historical record of each user's power generation and power consumption as power big data; The power prediction module preprocesses the power big data recorded in the previous step, selects appropriate features and creates large-scale samples; The power prediction module uses a radial basis function neural network to train a user power prediction model based on the training samples. Specifically, the power prediction module uses a radial basis function neural network to train a user power generation and power consumption prediction model based on the training samples. The power prediction module evaluates the model through the test samples to obtain the optimal user power prediction model. Specifically, the power prediction module evaluates the model through the test samples to obtain the optimal photovoltaic power generation prediction model and load demand prediction model; The photovoltaic power generation prediction model predicts the photovoltaic power generation of the user in the next trading time window, and the load demand prediction model predicts the load demand of the user in the next trading time window; an electricity collection module, used to establish an electricity purchase cost model and a seller benefit model, and calculate the electricity purchase cost of the system dispatcher and the benefit of the seller through the electricity purchase cost model and the seller benefit model, determine the electricity collection value collected from each electricity seller under the condition of minimizing the electricity purchase cost of the system dispatcher and maximizing the benefit of the seller, and perform electricity collection according to the determined electricity collection value collected by each electricity seller; and A power distribution module is used to establish a buyer benefit model, and calculate the sum of the buyer's benefits through the buyer benefit model, determine the power distribution value allocated to each power buyer on the condition of maximizing the sum of the buyer's benefits, and distribute power according to the power distribution value allocated to each power buyer to achieve a balance between power supply and demand in the microgrid; The user side is connected to the system dispatcher and includes a number of users, each of which is equipped with an electronic terminal, and a load and a power generation device connected to the system dispatcher through the electronic terminal; The system dispatcher performs power trading by controlling the power supply and demand balancing system.

2. The microgrid according to claim 1, It is characterized in that The power prediction module includes: A data loading submodule is used to obtain the power generation and load demand of each user in the microgrid in real time as power big data; A preprocessing submodule, used for preprocessing the power big data to select appropriate features and create training samples and test samples; A training submodule, used for training a user power prediction model based on the training samples; An evaluation submodule, used for evaluating the user power prediction model through the test sample to obtain an optimal user power prediction model; The prediction submodule is used to predict the power generation and / or load demand of the user in the next transaction time window through the optimal user power prediction model.

3. The microgrid according to claim 1, It is characterized in that The system also includes a prediction error processing module for performing compensation settlement on the transaction subject according to the prediction error.

4. The microgrid according to claim 3, It is characterized in that The prediction error processing module comprises: for performing compensation settlement according to the difference between the actual power sold by each of the power sellers and the power collected from the power sellers; and / or A buyer compensation settlement submodule is used for performing compensation settlement according to the difference between the actual purchased power of each power buyer and the power allocated to the buyer.

5. The microgrid according to any one of claims 1 to 4, It is characterized in that The microgrid trading system also includes a time window adjustment module, which is used to adjust the time window according to the prediction error and the system operation pressure.